How an AHU control strategy works, from start-up to shutdown

Air handling units are among the most complicated bits of plant a BMS looks after, and they’re behind a big share of comfort complaints and wasted energy. Let’s follow a typical AHU through a working day. Every site’s strategy is a bit different, so always work from the documented one, but most follow the pattern below.

What’s in the box

Follow the air through a typical supply and extract AHU and you’ll pass:

1. the fresh air damper
2. a pre-filter and a main filter
3. heat recovery, which might be a plate heat exchanger, a thermal wheel or a run-around coil
4. on some units, a frost or preheat coil
5. the heating coil (LTHW or electric)
6. the cooling coil (chilled water or DX), with its drain tray
7. the supply fan, usually EC or on a variable speed drive these days

The extract side has its own fan, filter and exhaust damper.

For sensors, expect supply, extract and outside air temperatures, differential pressures across the filters and fans, and sometimes humidity, CO₂ and duct static pressure.

Turning it on

The AHU gets enabled by a time schedule, an occupancy signal, optimum start or a manual override. It must be stopped by the fire alarm, a frost thermostat trip or a critical fault, and those always win over any schedule.

A safe start-up goes roughly like this:

1. Open the dampers first. Wait for the end switches to confirm, or for a set time, so the fans aren’t starting against closed dampers.
2. Start the fans. Often extract first and then supply, or both together, ramping the speed up rather than slamming them to full.
3. Prove the airflow. A differential pressure switch or sensor across each fan has to confirm air is moving within a set time, typically a minute or two. If it doesn’t, the unit stops and raises a fan failure alarm.
4. Release temperature control only once airflow is proven.

On cold mornings, a lot of strategies open the heating valve before the dampers, so the first rush of freezing air doesn’t trip the frost stat. There’s a whole article on AHU frost protection.

Controlling the temperature

Most AHUs control either the supply air temperature directly, or the room or extract temperature by adjusting the supply air setpoint in a cascade. Cascade is common in offices. If the extract air is warmer than its setpoint, the supply setpoint gets pushed down, within limits, say 14 °C to 28 °C.

The golden rule is to use free energy first and never heat and cool at the same time. A typical sequence driven by one PID loop looks like this:

Demand	What happens
Lots of heating	Heat recovery flat out, then the heating valve opens
A little heating	Heat recovery modulates, heating valve stays shut
Dead band	No heating or cooling
A little cooling	Heat recovery bypassed if the outside air is cooler than the extract
Lots of cooling	Cooling valve opens

That dead band in the middle matters. Without it, the loop flicks back and forth between heating and cooling. Overlapping sequences are one of the most common reasons you’ll find an AHU heating and cooling at the same time. If the loops themselves need sorting out, see PID loops in HVAC.

Fan speed

How the fans are controlled depends on the system:

• Constant volume. The fans run at a fixed speed set during commissioning.
• VAV. The supply fan holds a static pressure setpoint in the duct. Pulling that setpoint down when most VAV boxes are nearly shut saves a fair bit of fan energy.
• Demand-controlled. Fan speed follows CO₂ or occupancy, giving more air when spaces are busy and less when they’re empty.

The extract fan usually tracks the supply fan with an offset, keeping the building slightly positive or negative as the designer intended.

Humidity, where it matters

Some buildings, such as museums, labs and parts of hospitals, need humidity control too. The AHU humidifies with steam or adiabatic humidifiers, with a high-limit sensor downstream to stop the duct getting soaked. It dehumidifies by overcooling the air on the cooling coil to condense moisture out, then reheating it back to the supply setpoint.

That’s expensive, so it’s worth checking the humidity limits really are what the space needs.

Night purge

In summer, a night purge runs the AHU on cool outside air overnight to flush the day’s heat out of the building’s structure, so there’s less cooling to do the next morning. It typically runs when the outside air is cooler than inside by a decent margin, the building is warm enough to need it, and it isn’t so cold outside that people arrive to a chilly office.

Alarms and safeties

The usual list:

• fan failure (airflow not proven, or lost)
• dirty filters, from filter DP above its limit for a set time
• frost thermostat trip
• fire alarm stop, or smoke control mode where designed (this should be hardwired, not just a BMS point)
• supply temperature too high or too low
• condensate high level, where there’s a sensor

Shutting down

On a normal stop, the heating and cooling go back to their off positions. In cold weather, though, the heating valve often keeps a minimum opening to protect the coil. The fans stop, sometimes with a short run-on if there’s an electric heater. Then the dampers close, usually on spring return, so cold air can’t drift through the stopped unit overnight.

Where AHUs usually go wrong

When an AHU is misbehaving, these are worth checking first:

• heating and cooling running together because the sequences overlap
• a heat recovery bypass stuck in the wrong position, so the unit is heating outside air that’s actually warmer than the extract
• dampers that don’t close at night, which wastes heat and risks frozen coils
• a supply air sensor mounted too close to a coil
• fan proving failing because a DP tube has split or blocked
• the unit running 24/7 because someone added an override and forgot about it

When an AHU is playing up, a decent trend will usually show you which stage of this sequence it’s falling over at.